TLH01 commited on
Commit
84be09d
·
verified ·
1 Parent(s): 04c9cfa

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +16 -16
app.py CHANGED
@@ -2,7 +2,6 @@ import streamlit as st
2
  from transformers import pipeline
3
  from PIL import Image
4
  import numpy as np
5
- import io
6
 
7
  # Stage 1: Image to Text (Captioning)
8
  @st.cache_resource
@@ -17,19 +16,18 @@ def generate_caption(image):
17
  # Stage 2: Text to Story (Children-friendly)
18
  @st.cache_resource
19
  def load_story_generator():
20
- return pipeline("text2text-generation", model="mrm8488/t5-base-finetuned-common_gen", max_length=100)
21
 
22
  def text2story(description):
23
  story_gen = load_story_generator()
24
- prompt = f"generate a short story about {description}"
25
  story = story_gen(prompt)[0]['generated_text']
26
  return story
27
 
28
-
29
- # Stage 3: Story to Speech (Lightweight & Compatible)
30
  @st.cache_resource
31
  def load_tts():
32
- return pipeline("text-to-speech", model="espnet/kan-bayashi_ljspeech_vits", framework="espnet")
33
 
34
  def story_to_audio(story_text):
35
  tts = load_tts()
@@ -42,6 +40,7 @@ def story_to_audio(story_text):
42
  def main():
43
  st.set_page_config(page_title="Kids Story Creator", layout="centered")
44
  st.title("🧒 Kids Story Creator 📖")
 
45
 
46
  uploaded_image = st.file_uploader("Upload an image (jpg/jpeg/png):", type=["jpg", "jpeg", "png"])
47
 
@@ -49,18 +48,19 @@ def main():
49
  image = Image.open(uploaded_image).convert("RGB")
50
  st.image(image, caption="Uploaded Image", use_column_width=True)
51
 
52
- st.markdown("### Step 1: Generating Description...")
53
- caption = generate_caption(image)
54
- st.success(f"Caption: {caption}")
 
55
 
56
- st.markdown("### Step 2: Generating Story from Caption...")
57
- story = text2story(caption)
58
- st.success("Here's your story:")
59
- st.write(story)
60
 
61
- st.markdown("### Step 3: Converting Story to Audio...")
62
- audio, sample_rate = story_to_audio(story)
63
- st.audio(audio, format="audio/wav", sample_rate=sample_rate)
64
 
65
  if __name__ == "__main__":
66
  main()
 
2
  from transformers import pipeline
3
  from PIL import Image
4
  import numpy as np
 
5
 
6
  # Stage 1: Image to Text (Captioning)
7
  @st.cache_resource
 
16
  # Stage 2: Text to Story (Children-friendly)
17
  @st.cache_resource
18
  def load_story_generator():
19
+ return pipeline("text2text-generation", model="google/flan-t5-small", max_length=100)
20
 
21
  def text2story(description):
22
  story_gen = load_story_generator()
23
+ prompt = f"Generate a short children's story about {description}."
24
  story = story_gen(prompt)[0]['generated_text']
25
  return story
26
 
27
+ # Stage 3: Story to Audio (Lightweight & Compatible)
 
28
  @st.cache_resource
29
  def load_tts():
30
+ return pipeline("text-to-speech", model="coqui/tts-en-simply-tts")
31
 
32
  def story_to_audio(story_text):
33
  tts = load_tts()
 
40
  def main():
41
  st.set_page_config(page_title="Kids Story Creator", layout="centered")
42
  st.title("🧒 Kids Story Creator 📖")
43
+ st.write("Upload a picture and let us create a short story with voice for children aged 3–10!")
44
 
45
  uploaded_image = st.file_uploader("Upload an image (jpg/jpeg/png):", type=["jpg", "jpeg", "png"])
46
 
 
48
  image = Image.open(uploaded_image).convert("RGB")
49
  st.image(image, caption="Uploaded Image", use_column_width=True)
50
 
51
+ if st.button("Generate Story"):
52
+ with st.spinner("Step 1: Generating image description..."):
53
+ caption = generate_caption(image)
54
+ st.write(f"**Caption:** {caption}")
55
 
56
+ with st.spinner("Step 2: Creating children's story..."):
57
+ story = text2story(caption)
58
+ st.write("**Generated Story:**")
59
+ st.write(story)
60
 
61
+ with st.spinner("Step 3: Generating audio..."):
62
+ audio, sample_rate = story_to_audio(story)
63
+ st.audio(audio, format="audio/wav", sample_rate=sample_rate)
64
 
65
  if __name__ == "__main__":
66
  main()